Modern analytical methods such as SIMS and ESCA provide researchers witha vast amount of data in a short time. Traditional data reduction methods are not only inefficient but also make it difficult to identify patterns in and relationships between variables. Multivariate statistical methods can permit a more efficient use of the data collected, as they provide tools to assist in the optimization, interpretation, and extension of surface analytical techniques. Areas of application being investigated include: determination of spectral similarity applied to ESCA and TOF SIMS; quantitation of two overlapped components; computer spectral recognition; and correlation of surface properties with biological responses.